Abstract:

The Danjiangkou reservoir watershed is the source region of middle route of the SouthtoNorth Water Transfer Project.However,agricultural nonpoint source pollution is threatening the security of water environment in the watershed.Based on Geographic Information System (GIS) and Digital Elevation Model (DEM),the database of agricultural nonpoint source pollution for the reservoir watershed was established.According to the filed measured data,Soil and Water Assessment Tools (SWAT) were adjusted to analyze the temporalspatial distribution patterns of nonpoint source pollution load in the watershed.The agricultural nonpoint source pollution loads per unit area of major landuse types were also calculated.Results showed that there was a greater temporal difference in the output of agricultural nonpoint source pollution within the year in the watershed (p<001).The load of agricultural nonpoint source was strongly correlated with rainfall (rN=0896，rP=0920,p=001),sediment storage,ammonia nitrogen,total nitrogen and total phosphorus in the rainy period of June,July,August,accounted for 576%,64%,501%,508% and 473% of total year.The loss amount of soil nutrient such as nitrogen and phosphorus showed a great spatial variability in the watershed (p<001).Both farmland and bare land had a high contribution to the nonpoint source pollution load,whereas the contribution of forest to the pollution load was low.Its sediment load per unit area,nitrogen adsorption,dissolved nitrogen,phosphorus sorption and dissolved phosphorus were 715 t/hm2,892 kg/hm2,835 kg/hm2,807 kg/hm2,006 kg/hm2,respectively.This suggested that the agricultural nonpoint source pollution per unit area under different types of land use was obviously different.Under the simulated scenarios including different fertilizing amount and soil and water conservation measures,the reduction of fertilizing amount could effectively decline the pollution load of total nitrogen and total phosphorus,and returning farmland to forests or grasslands could notably reduce the output of sediment from farmland to river as well as the load of agricultural nonpoint source pollution in the watershed